Search

dUTP nucleotidohydrolase (dUTPase) catalyzes the hydrolysis of dUTP to dUMP and pyrophosphate and is the central regulator of cellular dUTP pools. Nuclear (DUT-N) and mitochondrial (DUT-M) isoforms of the protein have been identified in humans and arise from the same gene by the alternative use of 5′ exons. Recently, it has been shown that these isoforms are aberrantly expressed in some cancers and overexpression of dUTPase in the nucleus is associated with resistance to chemotherapeutic agents that target thymidylate biosynthesis. In this study, we have examined the signals necessary for dUTPase isoform localization using green fluorescent protein fusion constructs. We report that the N-terminal 23 amino acids of DUT-N are required but not sufficient for complete nuclear localization. Within this region, we identified a small cluster of basic residues (K14R15R17) that resemble a classic monopartite nuclear localization signal (NLS). Mutation of these residues completely abolishes nuclear localization. In addition, phosphorylation of Ser11 near the putative NLS has no affect on DUT-N nuclear localization. Through deletion analysis we show improved sorting of DUT-N to the nucleus when most of the protein sequence is present. Therefore, we conclude that DUT-N may contain a complex NLS that is located throughout the entire protein.

There are at least four distinct families of enzymes that recognize and remove uracil from DNA. Family-3 (SMUG1) enzymes have recently been identified and have a preference for uracil in single-stranded DNA when assayed in vitro. Here we investigate the in vivo function of SMUG1 using the yeast Saccharomyces cerevisiae as a model system. These organisms lack a SMUG1 homologue and use a single enzyme, Ung1 to carry out uracil-repair. When a wild-type strain is treated with antifolate agents to induce uracil misincorporation into DNA, S-phase arrest and cellular toxicity occurs. The arrest is characteristic of checkpoint activation due to single-strand breaks caused by continuous uracil removal and self-defeating DNA repair. When uracil-DNA glycosylase is deleted (Δung1), cells continue through S-phase and arrest at G2/M, presumably due to the effects of stable uracil misincorporation in DNA. Pulsed field gel electrophoresis (PFGE) demonstrates that cells are able to complete DNA replication with uracil-substituted DNA and do not experience the extensive strand breakage attributed to uracil-DNA glycosylase-mediated repair. As a result, these cells experience early protection from antifolate-induced cytotoxicity. When either UNG1 or SMUG1 functions are reintroduced back into the null strain and then subjected to antifolate treatment, the cells revert back to the wild-type phenotype as shown by a restored sensitivity to drug and S-phase arrest. The arrest is accompanied by the accumulation of replication intermediates as determined by PFGE. Collectively, these data indicate that SMUG1 can act as a functional homolog of the family-1 uracil-DNA glycosylase enzymes.

The thymidylate synthase reaction remains an important target for widely used anticancer agents; however, the clinical utility of these drugs is limited by the occurrence of cellular resistance. Despite the considerable amount of information available regarding mechanisms of drug action, the relative significance of downstream events that result in lethality remains unclear. In this study, we have developed a model system using the budding yeast Saccharomyces cerevisiae to dissect the influence of dUMP misincorporation into DNA as a contributing mechanism of cytotoxicity induced by antifolate agents. The activities of dUTPase and uracil-DNA glycosylase, key enzymes in uracil-DNA metabolism, were diminished or augmented, and the manipulated strains were analyzed for biochemical endpoints of toxicity. Cells overexpressing dUTPase were protected from cytotoxicity by their ability to prevent dUTP pool expansion and were able to recover from an early S-phase checkpoint arrest. In contrast, depletion of dUTPase activity leads to the accumulation of dUTP pools and enhanced sensitivity to antifolates. These cells were also arrested in early S-phase and were unable to complete DNA replication after drug withdrawal, resulting in lethality. Inactivation of uracil base excision repair induced partial resistance to early cytotoxicity (within 10 h); however, lethality ultimately resulted at later time points (12-24 h), presumably because of the detrimental effects of stable uracil misincorporation. Although these cells were able to complete replication with uracil-substituted DNA, they arrested at the G(2)-M phase. This finding may represent a novel mechanism by which the G(2)-M checkpoint is signaled by the presence of uracil-substituted DNA. Together these data provide both genetic and biochemical evidence demonstrating that lethality from antifolates in yeast is primarily dependent on uracil misincorporation into DNA, and that uracil-independent mechanisms associated with dTTP depletion play a minor role. Our findings indicate that the relative expression levels of both dUTPase and uracil-DNA glycosylase can have great influence over the efficacy of thymidylate synthase-directed chemotherapy, thereby enhancing the candidacy of these proteins as prognostic markers and alternative targets for therapeutic development.